Eccentricity and Higher Modes

Eccentricity, higher harmonics, and spin precession all redistribute power beyond a nonprecessing quadrupole chirp. In short or weak signals, their detector-facing morphologies can overlap with each other and with noise transients. Higher modes are established ingredients for some asymmetric binary-black-hole analyses; event-level eccentricity claims are more model and data dependent. A reliable inference compares models capable of representing the main alternatives and asks which time–frequency features actually drive the odds.

Helpful background. Binary Parameters and Observables develops the event likelihood. Modes, Harmonics, Frames, and Conventions defines waveform modes; Eccentric Waveforms explains their construction.

Three distinct sources of structure

For a quasicircular, nonprecessing binary, define the complex strain by $h=h_+-ih_\times$ and write

$$h_+(t)-ih_\times(t) =\frac{1}{D_L} \sum_{\ell,m}H_{\ell m}(t;\lambda) {}_{-2}Y_{\ell m}(\iota,\varphi).$$

Here $t$ is detector time and $H_{\ell m}$ uses detector-frame masses, with the overall luminosity-distance scaling removed. Using source-frame modes instead would require redshifting their times and normalization consistently. The linked waveform chapter fixes the remaining frame and phase conventions. The dominant $(\ell,m)=(2,\pm2)$ modes are not the entire signal. Subdominant modes grow relatively important for unequal masses and away from face-on orientation. They have different angular patterns and frequency evolution, so they can improve mass-ratio, inclination, distance, and polarization inference.

Precession rotates a coprecessing mode pattern into the inertial frame, producing amplitude and phase modulations and mode mixing. Eccentricity adds radial oscillations, multiple orbital harmonics, and altered phase evolution; its numerical value must be defined at a reference frequency or separation and within a waveform convention.

These mechanisms are physically distinct but not necessarily statistically orthogonal. A few high-amplitude cycles can be fitted by several combinations of orientation, precession, eccentricity, and higher modes.

Why eccentricity is difficult

Radiation reaction tends to circularize isolated binaries before they enter the ground-based band. Residual eccentricity can nevertheless arise from dynamical capture, repeated interactions, triples, or other accelerated pathways. Its observation would be informative—but precisely for that reason the null alternatives must be strong.

For long inspirals, eccentric harmonics and secular phasing can accumulate distinguishable structure. For a very massive binary observed for only a few cycles, an “eccentric” fit may instead exploit merger morphology not shared consistently among waveform families. Comparing an eccentric, nonprecessing model only with a quasicircular, nonprecessing model is incomplete if precession is plausible.

The likelihood must also distinguish coherent morphology from glitches. A feature concentrated in one detector, one short time interval, or one PSD choice should be stress-tested with alternative data conditioning and explicit glitch models.

Results and limitations through 2026-08-18

Official catalog parameter estimation uses waveform sets chosen for each analysis domain and includes precession and higher modes for informative binary-black-hole events. The GWTC-5.0 methods and observations papers document waveform comparisons and event-specific systematic checks LVK 2026, methods §§4–5 and LVK 2026, observations §3.

Evidence line	Supported conclusion	Remaining ambiguity
Asymmetric events show improved fit or parameter stability with higher modes	Quadrupole-only models can be inadequate	Individual mode amplitudes are not always independently measured
Some short, massive events favor eccentric over restricted circular baselines	Eccentric morphology is a credible hypothesis	Precession, higher modes, and waveform-domain error can give comparable fits
Injection studies recover cross-confusion	The degeneracy is expected from observable morphology	They do not select the correct interpretation for real non-Gaussian data
Catalog analyses remain broadly compatible with quasicircular populations	Large eccentricity is not required for most analyzed events	A small or selection-suppressed eccentric subpopulation is not excluded

GW190521 remains a useful adversarial example. An eccentric analysis found preference over restricted circular baselines but explicitly demonstrated confusion with precession Romero-Shaw et al. 2020, “Analysis of GW190521” and “Injection studies”. A later time-domain dissection showed how precession support in a quasicircular interpretation depends on the balance between quiet premerger data and merger–ringdown power Miller et al. 2024, §§ II–IV. The strongest conclusion is competing morphology, not uniquely measured orbital eccentricity.

By contrast, higher-mode evidence in asymmetric events such as GW190412 is supported by coherent multipolar analyses and physically expected amplitude scaling LVK 2020, §§IV–VI. Even there, parameter shifts and mode evidence depend on waveform accuracy and orientation priors.

A discriminating analysis

An analysis claiming eccentricity or higher modes should include:

1. compatible quasicircular quadrupole, higher-mode, precessing, and eccentric waveform comparisons;
2. matched priors or a transparent account of prior-volume differences in evidence;
3. detector-by-detector time–frequency localization of the discriminating power;
4. PSD, calibration, data-segment, and glitch-model variations;
5. injections of each competing morphology into representative real data;
6. waveform cross-validation against numerical relativity in the posterior region;
7. selection estimates before turning event odds into a population fraction.

Bayes factors alone are not enough. One should identify a statistic or waveform feature—secular eccentric phasing, a consistent harmonic relation, precession sidebands, or a subdominant angular mode—that the alternatives fail to reproduce.

Population and formation boundaries

Searches are not equally sensitive to eccentric, precessing, higher-mode, and quasicircular signals. A catalog selected by quasicircular banks cannot yield an eccentric fraction from event posteriors alone; its missed-signal efficiency must be measured with appropriate injections. Marginal triggers and burst searches may recover different parts of the space, requiring a coherent union of pipelines.

Even a robust eccentric event does not uniquely imply a globular-cluster origin. Nuclear clusters, triples, active-galactic-nucleus disks, and close dynamical captures can overlap. A formation claim must compare joint predictions for eccentricity at a declared frequency, masses, spins, redshift, and rates in Black-Hole Astrophysics.

The interpretation could change after a superseding catalog, a broader eccentric search with calibrated selection, a waveform comparison that changes event odds, or an independent data-quality reanalysis.

Exercises

1. Orientation and higher modes. Why are higher modes often less visible for a nearly face-on binary?

Solution

At the north orbital pole, the spin-weighted harmonic with spin weight $-2$ vanishes unless $m=2$; at the south pole, only $m=-2$ survives. Thus commonly important modes such as $(3,3)$ and $(4,4)$ are suppressed face-on. This is not a statement that every $\ell\gt2$ mode vanishes: modes such as $(3,2)$ can survive on the axis, although they are often intrinsically weaker than $(2,2)$. Unequal masses can increase higher-mode amplitudes, but their observed contribution still depends on orientation.

2. An unfair Bayes factor. An eccentric model with broad eccentricity prior is compared with a circular model that omits precession and higher modes. What can a preference establish?

Solution

It establishes only that the eccentric family predicts the data better after the stated prior-volume averaging than that restricted circular family. It does not isolate eccentricity, because the improvement may come from morphology that precession or higher modes could also fit. Those alternatives, compatible priors, and data-quality models are required for physical attribution.

References

• LIGO Scientific Collaboration and Virgo Collaboration. “GW190412: Observation of a Binary-Black-Hole Coalescence with Asymmetric Masses.” Physical Review D 102 (2020): 043015. DOI. Open PDF.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration. “GWTC-5.0: Methods for Identifying and Characterizing Gravitational-Wave Transients.” arXiv:2605.27224 (2026). Open PDF.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration. “GWTC-5.0: Observations from the Second Part of the Fourth LIGO-Virgo-KAGRA Observing Run and Updates to the Gravitational-Wave Transient Catalog.” arXiv:2605.27225 (2026). Open PDF.
• Miller, Simona J., Maximiliano Isi, Katerina Chatziioannou, Vijay Varma, and Ilya Mandel. “GW190521: Tracing Imprints of Spin Precession on the Most Massive Black Hole Binary.” Physical Review D 109 (2024): 024024. DOI. Open PDF.
• Romero-Shaw, Isobel M., Paul D. Lasky, Eric Thrane, and Juan Calderón Bustillo. “GW190521: Orbital Eccentricity and Signatures of Dynamical Formation in a Binary Black Hole Merger Signal.” Astrophysical Journal Letters 903 (2020): L5. DOI. Open PDF.